Advances-in molecular biology that have enabled the development of high throughput technologies for assaying gene expression, very large numbers of single nucleotide polymorphisms (SNPs), proteins, metabolic profiles etc. have revolutionized our ability to understand the biological basis of complex human disorders. As a consequence, we have now developed gene expression profiles that can predict treatment outcomes, and identified SNPs that are reproducibly associated with complex human diseases, findings that had previously been all but intractable. But it is difficult to be satisfied with the progress we have made when there is still so much that we do not know or understand about how common disorders arise and develop. We believe that we can learn much more from the systematic organization of the totality of the information developed through genome-wide interrogation of gene expression and DNA polymorphism than we have yet appreciated. Thus we propose to focus our expertise in statistical methods and algorithms for mining the data generated from genome-wide platforms toward a better understanding of the molecular architecture of psychiatric phenotypes. We will achieve this through a dynamic process of data acquisition, integration, phenotype deconstruction, and the development of a database, software and associated web browser that should provide the next logical step in merging the information that has started to become available from large throughput genotyping , sequencing, comparative genomic hybridization, and expression technology. The database will contain detailed annotation of all known genetic variants (including general information on location, conservation and local recombination rates, population characteristics such as frequency and evidence for selection, as well as association data on clinical and expression phenotypes), and of all genes (including general characteristics of location and variants within, information on pathways associated to the gene, as well as known SNPs and phenotypes associated with the gene). This project will build on an existing effort at University of Chicago called SCAN (SNP and CNV Annotation Network, www.scandb.org).